Explanation/Reference:
A: Smart Scan Column Filtering
Exadata provides column filtering, also called column projection, for table scans. Only the columns requested are returned to the database server rather than all columns in a table. For example, when the following SQL is issued, only the employee_name and employee_number columns are returned from Exadata to the database kernel.
SELECT employee_name, employee_number FROM employee_table.
For tables with many columns, or columns containing LOBs (Large Objects), the I/O bandwidth saved can be very large. Using both predicate and column filtering dramatically improves performance and reduces I/ O bandwidth consumption. In addition, column filtering also applies to indexes, allowing for even faster query performance.
C: Smart Scan Predicate Filtering
Exadata enables predicate filtering for table scans. Only the rows requested are returned to the database server rather than all rows in a table. For example, when the following SQL is issued only rows where the employees' hire date is after the specified date are sent from Exadata to the database instance.
SELECT * FROM employee_table WHERE hire_date > '1-Jan-2003'.
This ability to return only relevant rows to the server greatly improves database performance. This performance enhancement also applies as queries become more complicated, so the same benefits also apply to complex queries, including those with subqueries.